AI workflows remind me of the early days of the iPhone. I had to figure out how to bring the Hulu website experience to a small touchscreen while people were still working out how to use one. It was exciting, and it took a lot of trial and error. AI asks for the same mindset, and Tagalong is where I am working it out.

I have a son with Angelman syndrome. Keeping records is a real part of his care, and Tagalong does exactly that, and only that. I felt it was small enough to make with AI, so I decided to build it for parents like me.

It started fast with Claude. I got a working prototype in a few hours. I gave Claude the design tokens up front, and built the components as the features needed them. Claude implemented my mocks and redlines in minutes. However, getting from there to a working app took much longer. There was a lot to hash out. The new AI workflow was a real shift in how I work. Its strength is speed. Its weakness is attention to detail. Balancing the two is how I got the outcome I wanted.